Browse Articles

Endogenous VEGF signaling acts as a guardian of human primed pluripotency

TinyML pipeline for efficient crack classification in UAV-based structural health inspections

Abstract Structural health monitoring (SHM) of civil, aerospace, and energy infrastructure increasingly relies on UAVs with vision sensors for efficient inspections. Crack classification is a central task, yet cloud-based inference introduces bandwidth, power, connectivity, and privacy challenges that limit its practicality. This study presents a fully self-contained Tiny Machine Learning (TinyML) pipeline for onboard crack classification on a milliwatt-level STM32H7 microcontroller. Using MobileNetV1x0.25 as the baseline, we systematically evaluate the full measurement pipeline, including image capture, preprocessing, and inference on a low-power embedded system. Two preprocessing strategies, a handcrafted sequence (grayscale, contrast, denoise, median, binarization) and a greedy algorithm-based composite method, are compared. Four compression techniques, namely post-training quantization (PTQ), quantization-aware training (QAT), pruning, and weight clustering, are assessed individually and in combination. The optimized pipeline achieves an F1-score of 0.938, an improvement of 11.4% over state-of-the-art deployments. At the same time, it requires only 2.9 MB RAM and 309 KB flash, with an end-to-end latency of 461.6 ms and an energy cost of 623.16 mJ per inference. On a DJI Mini 4 Pro UAV, continuous operation reduces flight time by just 1.31 minutes (4%), compared to 8 minutes (24%) when using Jetson-based platforms. Overall, this work delivers a reproducible benchmark for UAV-based SHM, demonstrating a practical balance of accuracy, resource efficiency, and energy consumption, and advancing the feasibility of on-device crack classification in highly resource-constrained environments.

mist: a hierarchical Bayesian framework for detecting differential DNA methylation dynamics in single-cell data

N1-Methylpseudouridine directly modulates translation dynamics

Cycloastragenol attenuates osteoarthritis by restoring chondrocyte senescence via the NRF2/NF-κB signaling axis

DEUP1 functions as a scaffold for basal foot integrity and planar polarity in multiciliated cells

Statistics reach a ‘crisis point’: nations struggle with a critical lack of data

Internal conversion dominates the excited state dynamics and fluorescence efficiency of mono-substituted TPE-BODIPY dyes

Brachiopod genome unveils the evolution of BMP signalling in bilaterian body patterning

Abstract The molecular control and ancestral state of dorsal–ventral patterning within spiralians remain unclear due to the remarkable diversity across species. Here we present a chromosome-level genome for the brachiopod Lingula anatina and apply functional transcriptomics to study dorsal–ventral patterning under BMP signalling control. We uncover asymmetrical activation of BMP signalling at the dorsal side of the gastrula, governed by ventral chordin expression and a ‘seesaw’ of BMP ligands. Using small-molecule drugs and recombinant proteins, we show that high BMP activity inhibits genes typically associated with neural patterning during gastrula and larval stages, similar to deuterostomes and non-spiralian protostomes. Our findings suggest deep conservation of this mechanism across all three major bilaterian clades, supported by striking similarities in BMP-regulated gene sets between brachiopods and Xenopus . We argue that the spiralian ancestor retained the ancestral bilaterian mechanism, although downstream network components have undergone developmental system drift.

Decoding real-world visual scenes from alpha and gamma band flicker evoked oscillations in human EEG

Abstract Current approaches to investigate the role of neural oscillations in natural scene processing have been limited to artificial stimuli and long data collection. We present a new way to decode real-world scenes participants are viewing from the steady-state visually evoked potentials (SSVEPs) evoked while wearing flickering LCD glasses. We discovered that SSVEP responses from real world scenes are surprisingly complex and have distinct waveform shapes: they differ markedly across scenes and participants but are consistent within individuals, even across multiple days. SSVEP shape varies greatly between stimuli, but is reliable, meaning that decoding works even with a single electrode. Decoding is highly accurate with 5–10 s of data and was still above chance level with less than a second of data. Decomposing the SSVEPs into frequency bands showed that the information about the visual scene is present across all of the harmonics of the flicker frequency: optimal decoding used the broadband signals, but with 40 Hz (gamma band) showing the highest amount of information after band-pass filters. These findings implicate a broad range of oscillations in encoding real-world scenes, with a particular importance for 40 Hz. The SSVEP’s temporal profile is a rich source of information for decoding.

Visualising reaction complexes in amine-based unloaded and CO2-loaded carbon capture solutions

Abstract In power generation and industries where CO 2 emissions are unavoidable, carbon capture, utilisation, and storage is an important tool to offset climate change. Many carbon capture agents are blends of aqueous amines, which absorb CO 2 and are then thermally regenerated. The physical interactions between solutes play a crucial role in their reactivity and energy requirements for regeneration. Atomically resolved, experimentally derived information about the structure of these solutions, however, has yet to be reported. In this work, we report the structure of two model carbon capture solvents, aqueous sodium and potassium glycinate, in the unloaded and CO 2 -loaded state by performing structural refinement on H/D isotopically varied neutron diffraction data. This allows us to quantify the structure, frequency, and EPSR-derived pair interaction energetic stability of intermolecular interactions present. Such methodology can be readily applied to other carbon capture solutions, providing unparalleled insight and facilitating their large-scale modelling and rational design.

Genome-wide association study of resistance to taro leaf blight and yield traits in taro (Colocasia esculenta (L.) Schott)

Abstract Taro leaf blight (TLB), the most destructive disease of taro caused by the oomycete Phytophthora colocasiae . An effective strategy for controlling this disease is the development of resistant cultivars. In this study, a genome-wide association study (GWAS) was conducted on 265 taro genotypes to identify genomic regions and putative candidate genes associated with TLB resistance and yield traits. The field experiments were conducted at the National Root Crops Research Institute (NRCRI), Umudike, in 2021/2022 and 2022/2023 planting seasons. TLB severity, plant height, vigor, number of suckers, corm weight, cormel weight, and total tuber weight were evaluated. A genome-wide association mapping detected 18 single nucleotide polymorphisms (SNPs) markers significantly associated with TLB, number of suckers, corm weight, cormel weight, and total tuber weight. Gene annotation of significant SNP loci identified 6, 1, 2, 2, and 1 putative candidate genes associated with TLB resistance, number of suckers, corm weight, cormel weight, and total tuber weight respectively. Co-localization of a significant SNP 83978367 associated with corm weight and total tuber weight on chromosome 11 is possibly an indication of pleiotropic effects or the presence of closely linked genes controlling these traits. GWAS failed to detect significant associations in plant height and vigor. This study provides a basis for research on genetic control of TLB resistance, number of suckers, and yield-associated traits and thus provides resources for the ongoing effort to develop high-yielding, and disease resistant taro cultivars.

Thermal detection of single photons using Dirac fermions

Abstract Detecting single photons is a crucial process in quantum science, quantum networking, biology, and advanced imaging. To detect the small quantum of energy carried in a photon, conventional mechanisms rely on energy excitation across either a semiconductor bandgap or superconducting gap that hinders their applications to low-energy photons. Here, we detect single near-infrared photons using the thermal properties of Dirac fermions in graphene. By exploiting the extremely low heat capacity of Dirac electrons near its charge neutrality point, we observe a temperature rise up to  ~ 2 K using a hybrid Josephson junction. In this proof-of-principle experiment, we achieve an intrinsic quantum efficiency of 87% (75%) with dark count  < 1 per second (per week), reaching an effective noise equivalent power of 2 × 10 −22 W/ $$\sqrt{{{{\rm{Hz}}}}}$$ Hz . The highest operation temperature is 1.2 K. Our results highlight the potential of graphene bolometers for detecting lower-energy photons from the mid-IR to microwave regimes, opening pathways to study space science in far-infrared regime, to potential applications in dark matter searches, and to advance quantum technologies across a broader electromagnetic spectrum.

Human lateral occipital complex is invariant for position and size transformations at the single-neuron level

Design of miniprotein inhibitors targeting complement C9 to block membrane attack complex assembly

Abstract The abnormal formation of the membrane attack complex (MAC) is intrinsically linked to a range of acute and chronic immune diseases. The insertion of complement C9 into the membrane is the final step and kinetic bottleneck of MAC formation. However, research on blocking the MAC formation of C9 is currently limited. Given its broad, flat, and polar functional interface, complement C9 is a challenging target for rational design. Here, we utilize deep learning-based methods for protein scaffold generation, sequence design, and complex structure prediction to de novo design mini-protein inhibitors that specifically block the membrane insertion of soluble complement C9. The binding affinity of the mini-protein inhibitor is further optimized to 700 pM via partial diffusion. Design accuracy and binding specificity are verified through X-ray crystallography and biochemical studies. An in vivo acute hemolysis inhibition assay reveals that the C9 mini-protein inhibitors remain effective against hemolysis even 8 minutes after complement activation, outperforming the complement C5 inhibitor eculizumab. The de novo designed C9 mini-protein inhibitors can offer an optional therapeutic approach for the prevention and treatment of acute or chronic immune diseases associated with abnormal complement activation.

Ultra-bright supernova wobbles like a spinning top

Age-related alterations in trunk extensor force control during isometric and isokinetic contractions

Abstract Research on age-related changes in trunk extensor force control is currently limited, and the underlying neuromuscular mechanisms remain largely unexplored. To address this, we examined the relationship between oscillations in lumbar erector spinae (LES) activity and torque fluctuations in 20 young and 20 older adults during isometric and isokinetic (concentric) trunk extension contractions at 25% and 50% of maximal voluntary contraction (MVC). High-density surface electromyography (HDsEMG) signals were recorded bilaterally from the LES using 64-electrode grids. Torque steadiness was quantified using the coefficient of variation (CoV) of torque. Coherence analysis in the δ band (0–5 Hz) was applied between filtered interference HDsEMG and torque signals. Topographical maps were also generated to assess regional differences in HDsEMG-torque coherence. Older individuals exhibited greater torque CoV than young adults during both isometric (+ 23.03%, p  < 0.001) and isokinetic (+ 72.62%, p  < 0.001) contractions, with a larger between-group difference at 25% MVC for isokinetic contractions (Group × Torque interaction; p  = 0.007). At this intensity, the older group also showed reduced HDsEMG-torque coherence (Group × Torque interaction; p  = 0.004). During isometric contractions, coherence magnitude was similar across groups ( p  > 0.05), but older adults exhibited higher coherence in more cranial and medial LES regions ( p  = 0.005 and p  = 0.001, respectively). Older individuals exhibited the greatest impairment in force steadiness during low-intensity isokinetic contractions. Distinct neuromuscular patterns, possibly influencing force control, emerged depending on contraction type.

Kilogram-scale one-pot synthesis of multicomponent fullerene composites for efficient inverted perovskite solar cells

Integrated photonics enabling ultra-wideband fibre–wireless communication

Abstract Telecommunication systems are evolving towards ultrawide bandwidth and low latency, supporting wired and wireless links and their non-blocking interconnection 1 . However, a long-standing bandwidth mismatch between fibre communication and its wireless counterpart arises from fundamental disparities in signal architectures and hardware constraints 2,3 , which prevent high-speed and compatible transmission across the two domains. This challenge further complicates unified system design and hinders the realization of high-throughput-density, congestion-free fibre–wireless links under wideband-access scenarios 4 . Here we present an ultra-wideband (UWB) integrated photonics scheme that facilitates fibre–wireless communication over a shared-bandwidth infrastructure. Built on electro–optic (EO) and optic–electro (OE) conversions featuring 3-dB operational bandwidths exceeding 250 GHz and cross-architecture adaptability, our system demonstrates unprecedented data transmission capabilities in both wired and wireless links. Using the same set of devices and powered by the proposed complex bidirectional gated recurrent unit (complex-biGRU) algorithm, ultrahigh single-lane data rates of 512 Gbps for short-reach fibre and, for the first time to the authors’ knowledge, 400-Gbps high-speed wireless transmission have been achieved. Furthermore, high-density access is enabled by an all-optically assisted ultra-broadband wireless scheme. Real-time multichannel 8K video transmission is successfully demonstrated across 86 channels, seamlessly using a spectral range from 138 to 223 GHz. These findings in unified telecommunication development show the potential for the development of high-speed, densified and low-latency communication networks.

End-to-end example-based sim-to-real RL policy transfer based on neural stylisation with application to robotic cutting

Abstract Whereas reinforcement learning has been applied with success to a range of robotic control problems in complex, uncertain environments, reliance on extensive data - typically sourced from simulation environments - limits real-world deployment due to the domain gap between simulated and physical systems, coupled with limited real-world sample availability. We propose a novel method for sim-to-real transfer of reinforcement learning policies, based on a reinterpretation of neural style transfer from image processing to synthesise novel training data from unpaired unlabelled real world datasets. We employ a variational autoencoder to jointly learn self-supervised feature representations for style transfer and generate weakly paired source-target trajectories to improve physical realism of synthesised trajectories. We demonstrate the application of our approach based on the case study of robot cutting of unknown materials. Compared to baseline methods, including our previous work, CycleGAN, and conditional variational autoencoder-based time series translation, our approach achieves improved task completion time and behavioural stability with minimal real-world data. Our framework demonstrates robustness to geometric and material variation, and highlights the feasibility of policy adaptation in challenging contact-rich tasks where real-world reward information is unavailable.